Quote:
At any rate a mix of center frequencies(all at same amplitude) should be as much linear as possible.
That is some kind of dogma like the "less bands would be better"-theory.
IMO the problem is that the higher the complexity of a system is, the more difficult (
complicated) it is to control/tune it.
Linearity (as a concept/approach) is based on simplicity.
It is very easy to describe what linearity means (a regular progressing of constant intervalls).
IMO music and its spectral structure is very complex.
Of course it is based on regularities.
But it also contains somehow chaotic peaks based on the specific frequency range of different music and percussion instruments and their superimpositions.
That is why I believe that linearity for center frequencies in Multiband is not the best method.
Of course it is a lot of trial and error to manual edit STS files to find good/better multiband center frequencies, but I guess in the end you can achieve an overall superior processing comparing it with linear progressing settings.
So I think that I prefer Beta 1, because of its non-linearity.
Though I have tweaked it (Beta 1) a little bit.
In the past I had also used 8K (it was Bojcha who introduced that frequency first in a Stereo Tool preset !?), but now I prefer to use 7K instead, because I (
my ears) have the impression that 8K is one of these special (possibly most) superimposed peak points (voice-S-sounds + closed hihat).
That is why I prefer to shift the center of this band to 7k.